Just got back from a software course on a power monitoring meter that we have. I am wanting to set up alarms on THD. I asked someone in class what to set my alarms at. He said that he set his to alarm if THD Voltage goes above 5%. He said he doesn't worry about the THD current. Why not?
When would harmonics be a problem
- Thread starter JdoubleU
- Start date
- Status
mivey
Senior Member
Because at low load levels, the % current distortion can be huge, but as a percent of full load it would be relatively small.
The voltage will essentially tell you what is going on with the current.
winnie
Senior Member
- Location
- Springfield, MA, USA
- Occupation
- Electric motor research
Many loads are tremendously non-linear, and thus draw considerable harmonic current.
These loads consume/create harmonic current, and are not bothered in the least by the harmonic current.
Other things will be bothered by harmonic current; excessive heating in transformers and on the neutral, for example. These are things that care about _total_ loading, not % harmonics. They can carry 100% harmonic loading as long as the _total_ loading is low enough.
Finally there are sensitive loads which can be affected by harmonic _voltage_ on the supply. Harmonic _current_ being consumed from a supply will distort the output voltage of the supply, meaning that the non-linear loads drawing their harmonic current will result in a supply that shows harmonic distortion in its voltage. Since the supply voltage is relatively fixed, % harmonic voltage is a useful number if you are trying to protect such loads.
-Jon
These loads consume/create harmonic current, and are not bothered in the least by the harmonic current.
Other things will be bothered by harmonic current; excessive heating in transformers and on the neutral, for example. These are things that care about _total_ loading, not % harmonics. They can carry 100% harmonic loading as long as the _total_ loading is low enough.
Finally there are sensitive loads which can be affected by harmonic _voltage_ on the supply. Harmonic _current_ being consumed from a supply will distort the output voltage of the supply, meaning that the non-linear loads drawing their harmonic current will result in a supply that shows harmonic distortion in its voltage. Since the supply voltage is relatively fixed, % harmonic voltage is a useful number if you are trying to protect such loads.
-Jon
Would you be able to tell me how this happens. I some what understand what causes harmonics.
- Harmonics are normally created by non linear loads.
- The source of harmonics is a current source.
- By being a current source, harmonic current flows towards the lowest impedance on the network.
- At normal conditions, the lowest impedance is (must be) the generation side of the network (usually the feeding transformer)
- The voltage harmonics are being created by a voltage drop across the network impedance on the harmonic current flow path.
[/LIST
Practically, when the feeder's impedance at the harmonic frequency is low, even severe harmonic currents shouldn?t make much trouble to the facility itself.
The voltage harmonics level is the best indication if the impedance on the harmonic's way is low, so when voltage harmonics are OK, let's say THD is less than 3%, it is really insignificant what happens in the current.
In addition, since THD is only a measure of the harmonic value with reference to the fundamental frequency part. So, when the fundamental (first harmonic) value is low, means low load condition, the current THD value can go to the sky without any practical importance.
So, back to the original question:
Yes, voltage harmonics level even when presented by THD is much more important and indeed can indicate what happens in the current.
If you do want to trigger an alarm by harmonics troubles ? voltage is much better place.
PowerQualityDoctor
Senior Member
- Location
- Israel
pnelspec is right - voltage harmonics are of greater concern to identify big harmonics problem.
However, I would monitor current harmonics as well as they increase the losses significantly. The problem is with the THD level. When the load is low the THD can be very high without any real problem. Since that I recommend monitoring the TDD with a limit of 3% to 5%. Unfortunately, most meters do not measure the TDD.
If your meter does not have, I would have a limit of 15% to 20% on the main incoming, providing that the current is more than 10% of the service.
However, I would monitor current harmonics as well as they increase the losses significantly. The problem is with the THD level. When the load is low the THD can be very high without any real problem. Since that I recommend monitoring the TDD with a limit of 3% to 5%. Unfortunately, most meters do not measure the TDD.
If your meter does not have, I would have a limit of 15% to 20% on the main incoming, providing that the current is more than 10% of the service.
mivey
Senior Member
The previous two posts cover it but let's just think of an example. Suppose you have a very small load, say some electronic device drawing about 1 amp of mostly higher order currents on a 200 amp supply. Also suppose it is the only thing drawing current at the time of your measurement.
In this case, the harmonic distortion is huge because the current is practically all higher order currents. For the THDi, the numerator will be the square root of the sum of the squares of numbers much bigger than the denominator (the denominator is the fundamental current):
THDi = sqrt[sum{h=2 to infinity}(I_h)^2] / I_1
But, the impact on the system as a whole is minimal because it such a small portion of the system capacity. That is why we use TDDi which is the total demand distortion for current:
TDDi = sqrt[sum{h=2 to 50}(I_h)^2] / I_rated
or
TDDi = sqrt[sum{h=2 to 50}(I_h)^2] / I_max
...etc.
robbietan
Senior Member
- Location
- Antipolo City
voltage harmonics cause, among other things, overheating of conductors, transformers - which is why pocos usually limit these to 5%. a somewhat good approximation is 1%voltage THD = 8% current THD. but having high current THD is usually nothing to worry about - pocos look at current TDD (total demand distortion), the percentage of harmonic loads over the total loads.
harmonic load usually need a high current THD in order to operate - like those compact CFLs. however, even with their high current THD (usually 20%), there is no worry especially if you have a half-horsepower air conditioning unit in the same room. the load from the CFL is very small compared to that of the air conditioning unit.
Harmonics can be a serious problem for capacitors.
I have seen a fair number of failed power factor correction capacitors on sites that have a significant number of variable speed drives.
De-tuning reactors in series with the PFC would help to alleviate this problem but not everyone understands that.
I have seen a fair number of failed power factor correction capacitors on sites that have a significant number of variable speed drives.
De-tuning reactors in series with the PFC would help to alleviate this problem but not everyone understands that.
Harmonics are not exactly the problem for capacitors.
It is more like capacitors are the problem for harmonics. Capacitors in combination with network inductance can develop a parallel resonance circuit which can block harmonics on their way to the grid. This is the main problem which normally "kills" capacitors. BTW, in case of active parallel resonance capacitors should better die first, since if they don't the next will be a distribution transformer and the most sensitive loads.
Detuned reactors are not to protect capacitors! This is a very common mistake. Detuned reactors are to protect the network from being entering to a parallel resonance on the harmonic current frequency.
It is more like capacitors are the problem for harmonics. Capacitors in combination with network inductance can develop a parallel resonance circuit which can block harmonics on their way to the grid. This is the main problem which normally "kills" capacitors. BTW, in case of active parallel resonance capacitors should better die first, since if they don't the next will be a distribution transformer and the most sensitive loads.
Detuned reactors are not to protect capacitors! This is a very common mistake. Detuned reactors are to protect the network from being entering to a parallel resonance on the harmonic current frequency.
robbietan
Senior Member
- Location
- Antipolo City
as far as I know, capacitors are affected by high "current" harmonics, not voltage harmonics. the high frequency of the harmonic currents make the capacitor Xc lower, which makes it the point of least resistance in a circuit. Inductors XL increase as harmonics frequencies increase. The harmonic currents in a circuit then tend to go to the capacitors, which overheats and damages them
The low Xc on the harmonic frequency is a very good think, I wish it will be so simple and just a simple capacitors will become network least impedance forcing all current harmonics to go there. This would probably solve all harmonic problems for good. BTW, capacitors can easily withstand limited harmonic current and if properly designed, they do not being overheating just by some 30-50% of harmonic content. This is for example the case with passive filters where capacitors are designed to absorb harmonics.
However, the problem is when capacitors are forming a parallel resonance circuit with Xl of the grid and that what making the impedance on the harmonic way to reach the sky. This is one of the most dangerous situations on electrical network. And this is what "kills" capacitors and if not, everything else. Under parallel resonance condition the harmonic current thru the capacitor is almost uncontrolled, since the harmonic voltage is very high and as you've mentioned the Xc is very low.
Pol
- Status